You are on page 1of 11

Carnegie Mellon University

Research Showcase @ CMU


Human-Computer Interaction Institute School of Computer Science

2007

Research Through Design as a Method for


Interaction Design Research in HCI
John Zimmerman
Carnegie Mellon University

Jodi Forlizzi
Carnegie Mellon University

Shelley Evenson
Carnegie Mellon University

Follow this and additional works at: http://repository.cmu.edu/hcii


Part of the Graphics and Human Computer Interfaces Commons

This Conference Proceeding is brought to you for free and open access by the School of Computer Science at Research Showcase @ CMU. It has been
accepted for inclusion in Human-Computer Interaction Institute by an authorized administrator of Research Showcase @ CMU. For more information,
please contact research-showcase@andrew.cmu.edu.
Research through design as a method for interaction
design research in HCI
John Zimmerman, Jodi Forlizzi, Shelley Evenson
Human-Computer Interaction Institute and The School of Design
Carnegie Mellon University
{johnz, forlizzi, evenson}@andrew.cmu.edu
ABSTRACT effectively participate in HCI research.
For years the HCI community has struggled to integrate
design in research and practice. While design has gained a Through our inquiry we learned that many HCI researchers
strong foothold in practice, it has had much less impact on commonly view design as providing surface structure or
the HCI research community. In this paper we propose a decoration. In addition, we lack a unified vision of what
new model for interaction design research within HCI. design researchers can contribute to HCI research. This lack
Following a research through design approach, designers of a vision for interaction design research represents a lost
produce novel integrations of HCI research in an attempt to opportunity for the HCI research community to benefit
make the right thing: a product that transforms the world from the added perspective of design thinking in a
from its current state to a preferred state. This model allows collaborative research environment. The research
interaction designers to make research contributions based community has much to gain from an added design
on their strength in addressing under-constrained problems. perspective that takes a holistic approach to addressing
To formalize this model, we provide a set of four lenses for under-constrained problems.
evaluating the research contribution and a set of three To address this situation, this paper makes two
examples to illustrate the benefits of this type of research. contributions: (i) a model of interaction design research
Author Keywords designed to benefit the HCI research and practice
design, interaction design, interaction design research, HCI communities, and (ii) a set of criteria for evaluating the
research, research through design, wicked problems, design quality of an interaction design research contribution. The
theory, design method model is based on Fraylings research through design [14],
and it stresses how interaction designers can engage
ACM Classification Keywords wicked problems [21]. What is unique to this approach to
H5.2. User Interfaces: Theory and methods. interaction design research is that it stresses design artifacts
INTRODUCTION as outcomes that can transform the world from its current
In recent years we have both witnessed and participated in state to a preferred state. The artifacts produced in this type
the struggle as several academic institutions have attempted of research become design exemplars, providing an
to integrate design, with technology and behavioral science appropriate conduit for research findings to easily transfer
in support of HCI education and research. While there has to the HCI research and practice communities. While we in
been great excitement about the benefits integrating design no way intend for this to be the only type of research
can bring, we quickly realized that no agreed upon research contribution interaction designers can make, we view it as
model existed for interaction designers to make research an important contribution in that it allows designers to
contributions other than the development and evaluation of employ their strongest skills in making a research
new design methods. Over the last two years we have contribution and in that it fits well within the current
undertaken a research project to (i) understand the nature of collaborative and interdisciplinary structure of HCI
the relationship between interaction design and the HCI research.
research community, and (ii) to discover and invent Definitions
methods for interaction design researchers to more As we conducted this inquiry, we quickly realized that
within both the HCI and design communities there is an
inconsistent and confusing use of the following terms.
Therefore, below we provide a set of definitions for these
terms with respect to this paper.
Designer. Using such a generic term is a challenge at best.
At CHI 2006s SIG: The CHI Design Community, Bill
Buxton sarcastically claimed that if everyone is a designer
because they select their own clothes, then everyone is also
a mathematician, because we all count our change. His

1
comment captures what a loaded term designer is. Within synthesis of the findings from the literature and interviews,
the HCI community, it is quite common for people to use and the construction of a new model of design research;
the term design to mean HCI practice and to use the term iterative evaluations of this model with leading HCI
designer to mean an HCI practitioner. In this case a researchers and designers; and finally, a refinement that
designer might be an interaction designer, a usability produced the current model.
engineer, a software architect, a software developer, etc.
Literature review
However, in the design community, the term designer is
We reviewed the design research literature to understand
generally used to refer to someone who has had training or
historical and currently proposed models of design research
extensive practical experience in a discipline such as
and more specifically, interaction design research. In
architecture, product design, graphic design, or interaction addition, we reviewed literature from the HCI community
design. As we use the term designer in this paper, we are discussing the role of design.
following the convention of the design community.
CHI 2004 workshop
Design research. In the HCI community and in the design In 2004 we conducted a workshop at the CHI conference in
practice community, the term design research is generally Vienna, focusing on clarifying the relationship between
used to refer to the upfront research practitioners do to HCI and design. The workshop, had 22 participants from
ground, inform, and inspire their product development both academia and industry and from a range of
process. However, in the design research community, backgrounds including computer science, behavioral
including institutions such as the Design Research Society, science, and interaction design and explored two distinct
the term design research implies an inquiry focused on but complimentary tracks: (i) role of design in HCI
producing a contribution of knowledge. This paper follows education and role of HCI in design education, and (ii) the
the convention of the design researchers, and we intend the role of interaction design research in HCI. Outcomes from
term design research to mean an intention to produce this workshop helped frame our focus on the need to define
knowledge and not the work to more immediately inform models of design research in HCI and motivated us to
the development of a commercial product. engage the broader HCI practice and research community in
Design thinking. This term is often used to describe what a discussion of what these might be.
designers bring to problem solving and to rationalize why Interviews
designers need to be included in a project or process; We conducted semi-structured interviews with nine leading
however, it is rarely defined. In some respects its ambiguity academic HCI researchers accompanied by one of their
is part of its strength, allowing it to be the right thing at the graduate students. During the interview, we asked about
right time. In terms of this paper, we mean the application their ideas of what design is and what design research is in
of a design process that involves groundinginvestigation terms of HCI research. We chose to interview the leaders
to gain multiple perspectives on a problem; ideation with a graduate student for two reasons. First, we thought a
generation of many possible different solutions; iteration process of co-discovery would help us elicit better
cyclical process of refining concept with increasing fidelity; information during the interview. Second, we wanted to see
and reflection. if the students, who were much newer to HCI and were
In the next section, we provide an overview of our research being educated in a multidisciplinary environment that
and methodology in constructing this model. We highlight includes behavioral science, computer science, and
the findings from our literature review, and detail the interaction design, had a substantially different view of
evolving history of design in HCI and of interaction design design than their advisors, who had all been trained in a
research and its impact on the HCI community in order to single discipline.
situate our contribution within the frameworks of HCI We also interviewed six leading interaction designers.
research and design research. We then describe the model, Three held senior academic positions, and three held
detail how it produces knowledge, and discuss how it industry positions including head of design at a consumer
produces benefits for both the HCI practice and research electronics company, a design researcher at a well-known
communities. We formalize the model by describing four technology research company, and the principal of a design
lenses to evaluate the quality of an interaction design consultancy. In these interviews, in addition to collecting
research contribution. Finally, we illustrate how three information on the evolution of their career in HCI, we
examples of interaction design research can be evaluated by probed on the nature of the relationship between design and
the criteria described here. HCI and on what they saw as the important models of
METHODOLOGY design research with respect to the HCI research
Our methodology included a literature review focusing on community.
design in HCI and on models of design research; a Synthesis, analysis, and iterative modeling
workshop on the relationship between design and HCI; After generating a preliminary model, we iteratively
semi-structured interviews with leading HCI researchers evaluated the model through presentations and discussions.
and leading interaction designers in academia and industry; One included a large group of HCI researchers, none of
whom had training in design. In addition, we held four one- accurately modeled and cannot be addressed using the
on-one presentations of the model along with other research reductionist approaches of science and engineering [21].
models including Dick Buchanans model of design They argued that many problems can never be accurately
research [4] and Daniel Fallmans model of research- modeled, thus an engineering approach to addressing them
oriented design and design-oriented research [12,13]. These would fail. This work pointed to an opportunity for design
one-on-one interviews included a senior HCI practitioner, a research to provide complementary knowledge to the
leading design researcher in HCI, and two leading HCI contributions made by scientists and engineers through
researchers from industry. The one-on-one discussions methods unique to design and design processes.
allowed for more free-form feedback on our model and a
Design researchers describe their work as the study,
chance for the interviewee to participate in rapid redesign.
research, and investigation of the artificial made by human
The large discussion was particularly beneficial in that it
beings, and the way these activities have been directed
engaged the entire group in a discussion of what design
either in academic studies or manufacturing organizations.
research meant to their specific discipline within HCI and a
[3 p 16.] The focus of this work has been on a study of
discussion of what design does and should mean within the
design in order to improve the process and on the analysis
HCI research community.
of design artifacts in order to generate theories that unite
LITERATURE REVIEW related methods of addressing design challenges. In general,
Our literature review, meant to ground our inquiry, focused this design research scholarship has not focused on the
on the history of the emerging field of design research, the outcome (artifacts) of making as a design contribution.
role of design in HCI, and the role of the designed artifact.
The Role of Design in HCI
Design Research In the early days, the term design within the HCI
The emergence of design research as a separate activity community meant usability engineering: the process of
from design practice grew out of the need to formally modeling users and systems and specifying system behavior
address the increasing complexity of systems designers such that it fitted the users tasks, was efficient, easy to use
were being asked to create [3]. The increasing complexity and easy to learn. [26 p.1]. Over time, trained designers
of products such as battleships, airplanes, and rockets began working with software developers, bringing skills in
created a need for new design methods that were more visual hierarchy, navigation, color, and typography they had
predictable and more collaborative. The design methods developed designing printed artifacts. Jonas Lwgren
movement grew out of this need, and generated the first labeled the process they brought to interaction design as
cohort of design researchers focusing on the development creative design to distinguish it from the engineering
of knowledge instead of artifacts for consumption. approach [16]. In engineering design, developers created
Within the design research community, there has been an software to meet a specification, and in creative design,
designers continually reframed the problem, constantly
ongoing tension around the relationship between design and
questioning the underlying assumptions during the design
science [8]. Motivation for a scientific framing came from
process.
sources such as Buckminster Fullers call for a design
science revolution based on science, technology, and Daniel Fallmans work casts HCI as a design discipline
rationalism [8 p.50], and from Herbert Simons call for [12]. He describes the research performed by engineers and
the study of science of design to help more liberally educate behavioral scientists as design-oriented research.
scientists and engineers in his book Sciences of the Researchers engage in designing and making prototypes in
Artificial [23]. In this case the science can be a scientific order demonstrate a research contribution. In this case, the
study of how designers work or the use of scientific research community benefits from the processes of design
knowledge and methods in a rational practice of design [8]. and design thinking because they lead to better research
In adding to the research discussion of design methods, prototypes.
Donald Schn introduced the idea of design as a reflective Christopher Alexanders work on Pattern Languages
practice where designers reflect back on the actions taken in represents an example of how research performed by design
order to improve design methodology [22]. While this may researchers on design methods has had an impact on the
seem counter to the science of design, where the practice of HCI community. His work asks design researchers to
design is the focus of a scientific inquiry, several design examine the context, system of forces, and solutions used to
researchers have argued that reflective practice and a address repeated design problems in order to extract a set
science of design can co-exist in harmony [8, 5]. underlying design patterns, thereby producing a pattern
In reaction to the casting of design as a science and also in language [1]. The HCI community has embraced this
response to systems engineers inability to apply scientific approach to address design of web sites [24]. The method
methods to address social problems such as urban crime, turns the work of many designers addressing the same
Horst Rittel and Melvin Webber proposed the concept of a interaction problems into a discourse for the community,
Wicked Problem, a problem that because of the allowing interaction designers to more clearly observe the
conflicting perspectives of the stakeholders cannot be

3
formation of conventions as the technology matures and is is understood through reflective practice, intellectual
reinterpreted by users. apperception, and intentional choice. The practice of design
is framed as encompassing the real, the true, and the ideal;
The Artifact as a Part of Interaction Design Research
Daniel Fallman describes the HCI development process design research is framed as research on a condition that
used today as research-oriented design to describe the arises from a number of phenomena in combination, rather
research performed to influence the design of commercial than the study of a single phenomenon in isolation. Our
model of interaction design research in HCI attempts to
products [12]. Brenda Laurels book, Design Research:
formalize many of their ideas in a single method tailored to
Methods and Perspectives, also describes how interaction
fit within the context of the HCI research community. We
designers can perform research as they practice design to
better ground their process and to hopefully increase the do not view our model as the only way for interaction
chances for success of a product in the marketplace [15]. designers to perform research, but as one of many.
Through this process, HCI practitioners and interaction Our model of design research advances the work of the
designers work together as team members, keeping the design research community by expanding their focus on
needs of the user in focus for the entire development team. methods and analysis of artifacts to include making as a
While both represent a combination of research and method of inquiry in order to address wicked problems. Our
making, the focus is still on design as a practice and not as a model builds on the current relationship design has with the
research discipline that makes contributions of knowledge. HCI community by building on Alexanders pattern
At last years CHI conference a paper argued against a language model as a method of making research findings
actionable by the HCI practice community. Finally our
commonly held belief in the HCI research community that
model adds a counterpoint to critical designs focus on
design is a Black Art [25]. The authors argued instead
that interaction design performed in a research context design research in the role of critic, by creating a role for
employs a set of rational judgments. The case documented the design researcher to be an equal collaborator with HCI
in this paper places interaction design in the context of HCI engineering and behavioral science researchers.
research and interaction designers as collaborators with INTERVIEWS
researchers. However, in this specific case, the designers Interviews were conducted within and outside of the
work in service of research, with the goal of creating a discipline of interaction design to understand how
research prototype that more clearly communicates the collaborations between design and HCI evolve, and to
research contribution. We certainly see this type of iteratively test our model in progress.
collaboration as important to the ongoing relationship
Interaction designers on design and design research
between researchers and designers, but push for additional In our interviews with designers we probed on the value
collaborations where designers also participate in research design brings to HCI, and three main themes emerged.
and engage research questions specific to interaction
First, participants noted that interaction designers brought a
design.
process for engaging massively under-constrained problems
Critical design presents a model of interaction/product that were difficult for traditional engineering approaches to
design making as a model of research [9]. Unlike design address. Second, designers brought a process of integrating
practice, where the making focuses on making a ideas from art, design, science, and engineering, in an
commercially successful product, design researchers attempt to make aesthetically functional interfaces. One
engaged in critical design create artifacts intended to be described this process as similar to composing music or
carefully crafted questions. These artifacts stimulate conducting a symphony, where the job is to bring out the
discourse around a topic by challenging the status quo and richness in a range of voices to make a singular thing.
by placing the design researcher in the role of a critic. The Third, designers brought empathy for users as a part of the
Drift Table offers a well known example of critical design process. In addition to considering their needs and desires
in HCI, where the design of an interactive table that has no from an external-observers perspective, designers worked
intended task for users to perform raises the issue of the to also embody the people they made things for.
communitys possibly too narrow focus on successful
The designers we spoke with described their early days of
completion of tasks as a core metric of evaluation and collaborating with HCI and software developers, where
product success [10]. they were often brought in at the end of the process and
Finally, in their book, The Design Way, Harold Nelson and asked to make the interface pretty. In attempting to
Erik Stolterman frame interaction designand more improve the designs, they were often frustrated that the
generally the practice of designas a broad culture of suggestions they made, which often seemed obvious design
inquiry and action. They claim that rather than focusing on improvements, could not be made because they came too
problem solving to avoid undesirable states, designers work late in the development process. However, over time,
to frame problems in terms of intentional actions that lead designers moved from a consultant role at the end of a
to a desirable and appropriate state of reality. Design is project to team members working throughout the software
viewed as a unique way to look at the human condition, and design and development cycle.
Our interviewees suggested that recently, the shift from a above in a format that complements current methods of
more narrow focus on work to a broader view of interaction HCI research. It follows from Christopher Fraylings
in peoples lives has increased the role of designers in HCI concept of conducting research through design [14] where
research and practice. Industrial designers, communication design researchers focus on making the right thing; artifacts
designers, and newly minted interaction designers all began intended to transform the world from the current state to a
to play more important roles in the invention and preferred state.
development of radically new artifacts meant to address a
In our model (Figure 1), interaction design researchers
broad set of problems and opportunities.
engage wicked problems found in HCI. Examples of
Our interviews with leading interaction designers showed wicked problems include: (1) The design of smart home
that while they have strong agreement about the role design services for families where parents address the paradox of
plays and the benefits it brings to HCI practice, designers wanting to care and protect their children while also
lack clarity on what design research is or should be with wanting to make them independent and children face the
respect to HCI research. In all cases our questions about paradox of desiring the comfort and security their home and
design research performed outside of a specific design case family provide while also wanting to step out and discover
caught our participants off guard. However, through and invent who they are and might be. (2) The role of
probing on this issue, three roles for design to play in ubiquitous, assistive technology in aiding an elderly
research emerged: (i) design researcher in service of a population to age in place in their own homes. It is
research communityworking to help researchers ground wicked in that the stakeholders have conflicting goals
and frame problems and communicate the impact; (ii) including adult children who often want their parents out of
design researcher as critic of the HCI communitymaking the home in an environment that can better ensure their
artifacts that stimulate discussion of critical issues; and (iii) safety, and elder parents who have huge identity
design researcher as pattern finder, finding patterns that investments in their homes, and desire to remain, even
lead to pattern languages. when doing so creates tremendous social isolation.
One challenge a few interviewees noted for designers Using our model, interaction design researchers integrate
participating in research comes from the consultancy model the true knowledge (the models and theories from the
that drives most design work. Since the majority of design behavioral scientist) with the how knowledge (the technical
research is paid for by the development industry, it is opportunities demonstrated by engineers). Design
unlikely that this information, which provides a significant researchers ground their explorations in real knowledge
competitive advantage, would be openly shared. produced by anthropologists and by design researchers
performing the upfront research for a design project.
HCI researchers on design and design research
Through an active process of ideating, iterating, and
While interaction designers could articulate their role
within an HCI team, researchers were far less articulate critiquing potential solutions, design researchers continually
about the role of interaction design. In our interviews with reframe the problem as they attempt to make the right
leading HCI researchers, we heard views of design as the thing. The final output of this activity is a concrete problem
discovery of mental models, a discipline focused on the framing and articulation of the preferred state, and a series
whole instead of the parts, and desire to understand of artifactsmodels, prototypes, products, and
users. However, the dominant view was that designers documentation of the design process.
focused solely on the surface structure, or the visual This research through design approach produces several
aesthetics of software and hardware artifacts. This idea of beneficial contributions for the HCI community. First,
design and designers as having a focus on decoration is a design researchers identify opportunities for new
commonly held belief of design by most people [5]. technology or for advancements of current technology that
will have significant impact on the world. This type of
When asked about what design research is and what design
design research provides research engineers with inspiration
researchers do, the HCI researchers we interviewed had no
concrete ideas. This is not surprising given the lack of and motivation for what they might build. Design
clarity within the interaction design community on what researchers also undertake problem framing that helps
design is. Instead, a common theme we heard was that it identify important gaps in behavioral theory and models. In
was up to the interaction designers working in research to evaluating the performance and effect of the artifact
invent what design research should be within the context of situated in the world, design researchers can both discover
HCI. unanticipated effects and provide a template for bridging
the general aspects of the theory to a specific problem
A MODEL OF INTERACTION DESIGN RESEARCH space, context of use, and set of target users.
WITHIN HCI RESEARCH
Our research model attempts to unite and advance the
findings from the literature review and interviews described

5
Figure 2. An illustration of the pathways and deliverables between and among Interaction Design Researchers and
HCI Researchers. The model emphasizes the production of artifacts as vehicles for embodying what ought to be and
that influence both the research and practice communities.

Second, interaction design researchers create artifacts contribution of a whole, engineers first develop a
provide concrete embodiments of theory and technical specification of what they need to make to meet a specific
opportunities. These design exemplars then become an need. Next, they take a research focus asking questions
appropriate conduit for the transfer of HCI research to the such as can this be built? Is there a better way to build this?
practice community. Design artifacts are the currency of
In proposing a model of design research with a focus on the
design communication. In education they are the content
production of artifacts, we build on Nigel Crosss idea that
that teachers use to help design students understand what
design knowledge resides in the product [7]. The artifact
design is and how the activity can be done [7]. In research,
reflects a specific framing of the problem, and situates itself
they describe a vision of a preferred state, increasing the
in a constellation of other research artifacts that take on
chance for knowledge transfer to the research, practice, and
similar framings or use radically different framings to
education communities. Through exposure to the ideas in
address the same problem. These research artifacts provide
the artifacts, the practice community can more easily
the catalyst and subject matter for discourse in the
observe the value of different theories, models, and
community, with each new artifact continuing the
technology, and this can motivate them to follow the
conversation. When several related research artifacts have
threads back to the original research that might most impact
been created, then researchers can use more traditional
their work.
design research methods to analysis the artifacts and search
Third, use of this model results in holistic research for similar approaches designers have taken in addressing
contribution that reveals the framing of the problem and the common problems. The artifacts made through design
balance the researchers have made between the intersecting research have the potential to become pre-patterns [6] from
and conflicting perspectives. The idea of contributing a which design patterns [1] can begin to emerge.
whole closely resembles the work of systems engineers in
Our model departs from the roles of the design researcher
HCI who focus on building whole systems. The Aware
discovered in our literature review and interviews: (i)
Home constructed at Georgia Tech provides a good
design researcher as member of design practice team doing
example [11]. In this case the novelty was not in the
upfront project research; (ii) traditional knowledge
construction of the individual elements, but in the
producing design researcher studying design process and
integration of many technical research contributions from a
analyzing artifacts to discover patterns; (iii) design
variety of disciplines, into a single working system. The
researcher as critic; and (iv) design research as framing and
difference between this type of contribution and the design
communication consultant in service of other researchers.
research contributions we propose involve both the intent
Using our model, design researchers work in a way very
and the process of the research. In making a technical
similar to design practitioners, applying their strength at community, there is no agreed upon standard of what
addressing under constrained problems. This is in no way research through design means nor what a high quality
intended to discount the other design research roles that contribution should be. To help to formalize this research
have already had significant impact on HCI. Instead, we are method, we propose a set of criteria, or four lenses for
proposing an additional model of design research that evaluating an interaction design research contribution:
seems particularly suited for interaction design researchers process, invention, relevance, and extensibility.
working in HCI research and allows design researchers to
Process: One of the critical elements for judging the quality
work more as a collaborative equal with other HCI
of an interaction design research contribution is the process.
researchers. An obvious criticism of this model is how in its
Like anthropologists making contributions in this science-
use design researchers can distinguish their contributions as
dominated domain, there is no expectation that reproducing
research and not as practice. This is a concern raised by
the process will produce the same results. Instead, part of
Nigel Cross, who cannot consider normal works of practice
the judgment of the work examines the rigor applied to the
to be regarded as research contributions [7].
methods and the rationale for the selection of specific
We differentiate research artifacts from design practice methods. In documenting their contributions, interaction
artifacts in two important ways. First, the intent going into design researchers must provide enough detail that the
the research is to produce knowledge for the research and process they employed can be reproduced. In addition, they
practice communities, not to make a commercially viable must provide a rationale for their selection of the specific
product. To this end, we expect research projects that take methods they employed.
this research through design approach will ignore or de-
Invention: The interaction design research contribution
emphasize perspectives in framing the problem, such as the
must constitute a significant invention. Interaction design
detailed economics associated with manufacturability and
researchers must demonstrate that they have produced a
distribution, the integration of the product into a product
novel integration of various subject matters to address a
line, the effect of the product on a companys identity, etc.
specific situation. In doing so, an extensive literature review
In this way design researchers focus on making the right
must be performed that situates the work and details the
things, while design practitioners focus on making
aspects that demonstrate how their contribution advances
commercially successful things.
the current state of the art in the research community. In
Second, research contributions should be artifacts that addition, in articulating the integration as invention,
demonstrate significant invention. The contributions should interaction designers must detail how advances in
be novel integrations of theory, technology, user need, and technology could result in a significant advancement. It is
context; not just refinements of products that already exist in the articulation of the invention that the detail about the
in the research literature or commercial markets. The technical opportunities is communicated to the engineers in
contribution must demonstrate a significant advance the HCI research community, providing them with guidance
through the integration. This aspect of a design research on what to build.
contribution makes particular sense in the interaction design
Relevance: Scientific research has a focus on validity. In
space of HCI. Meteoric technological advances in hardware
engineering, this often means a demonstration of the
and software drive an aggressive invention of novel
performance increase or the function of their contribution.
products in HCI and interaction design domains that are not
In behavioral science, validity means an experiment that
as aggressively experienced by other design domains.
disproves the null hypothesis. In both cases, the work must
While product designers might find themselves redesigning
be documented in such a way that peers can reproduce the
office furniture to meet the changing needs of work,
results. As mentioned above, this does not make sense to
interaction designers more often find themselves tasked
have as a requirement for a research through design
with inventing whole new product categories.
approach. There can be no expectation that two designers
Our model of design research allows interaction design given the same problem, or even the same problem framing,
researchers to do what designers do best: to study the world will produce identical or even similar artifacts. Instead of
and then to make things intended to affect change. Our validity, the benchmark for interaction design research
model provides a new channel for the power of design should be relevance. This constitutes a shift from what is
thinking, desired by many disciplines, to be unleashed as in truethe focus of behavioral scientists, to what is realthe
a research context. Design researchers can contribute from focus of anthropologists. However, in addition to framing
a position of strength, instead of aping the methods of other the work within the real world, interaction design
disciplines as a means of justifying their research researchers must also articulate the preferred state their
contribution. design attempts to achieve and provide support for why the
community should consider this state to be preferred.
CRITERIA FOR EVALUATING INTERACTION DESIGN
RESEARCH WITHIN HCI Today, many design research contributions claiming to
Many design researchers have made contributions using a follow a research through design approach neglect to cast
research through design approach. While the idea is not the work in terms of relevance. The design researchers
new, within the HCI and interaction design research

7
follow a design process, but the motivation for their work,
the detail on current situation, and on the preferred state are
missing. Without this critical component, a research
through design approach appears to be self-indulgent, and
personal exploration that informs the researcher but makes
no promise to impact the world.
Extensibility: The final criterion for judging successful
design research is extensibility. Extensibility is defined as
the ability to build on the resulting outcomes of the
interaction design research: either employing the process in
a future design problem, or understanding and leveraging
the knowledge created by the resulting artifacts.
Extensibility means that the design research has been
described and documented in a way that the community can
leverage the knowledge derived from the work. Figure 3. Xerox prototype machine.
EXAMPLES OF INTERACTION DESIGN RESEARCH In terms of invention, this work demonstrated an integration
WITHIN HCI of the latest cognitive research on how people learn to
In order to demonstrate how the lenses might work for interact with systems. In terms of relevance, it connected
evaluating the quality of interaction design research in HCI, with the increasing need in the work place to empower
we provide examples of three interaction design cases that workers to take more ownership and responsibility for the
help illustrate different aspects of this model. individual documents they were working on. One of the
XEROX reprographics most valuable contributions was extensibility. The Xerox
FitchRichardsonSmiths work with Xerox in the early guidelines and rationale document communicated reusable
1980s on the interaction design of reprographics machines information for extensions in design [27] but the machines
provides an early example of research through design that themselves became objects that could be read by other
produced design exemplars and a design language (an designers outside of Xerox. This worked to transfer the
intentional pattern language) that can still be seen today in knowledge to the practice community. Today, elements and
the interaction and behavior of copiers and printers. The resources from this interaction design research project can
design process was documented in an extended rationale, still be seen in almost every copier and printer.
called Principles for Constructing Communicative Objects Philips vision of the future
and Object Systems for Interactive Dialogs, and detailed the In 1995, Philips Designs Vision of the Future project
design and rationale for every element of a machine to explored possibilities for life and technology in the near
support positive interaction [28]. future. Using a rigorous design process documented in the
Prior to this work, reprographics machines used in offices book Vision of the Future [19], this project examined how
generally had a key operator: a trained technician that held advances in technology would change family life along
the key to operate, maintain, and repair the machine. Design with other aspects of society. Multidisciplinary teams were
researchers working on this project reframed the problem brought together to propose directions for new products and
from making a machine that was easier for a key-operator services in four different domains of life: personal,
to maintain to making a machine that any office worker domestic, public, and mobile. In terms of relevance, the
could walk up to, use, and fix if it had a paper jam. The work detailed how changes to traditional forms and
prototypes produced (Figure 3) illustrated the idea that behaviors of technical products technology could allow
people could learn to operate the machines as they used products to more easily integrate into the social life of
themrather than being trained, which was unheard of in people outside of work environments. For example, Figure
the industry at that time. The design language included the 4 shows a mobile communication device housed in an
use of green on the copy button and on the edge of the glass aesthetic form not unlike a flower vase. The novel designs
panel to indicate points of entry, and the color blue to clearly had the intention of improving the quality of
indicate where users should interact with paper. Lighter peoples lives and provided a view of a preferred state. The
shades indicated areas of frequent interaction and darker work and the design process have also proven to be
shades indicated areas with less frequent use. Texture extensible, as numerous undergraduate and graduate
indicated specific touch points. Finally, the prototypes programs in interaction design have imitated the project.
provided concrete illustrations of how to provide
instructions at the point of need. Evaluations of the
prototypes revealed a shift in work practice that came about
as a result of the new way of interacting with the machines
[20].
world. Fourth, it allows interaction designers to make
research contributions that take advantage of the real skill
designers possessreframing problems through a process
of making the right thing. Fifth, it motivates the HCI
community to discuss preferred states and to reflect on the
potential impacts research might have on the world.
We hope that in proposing this model, we can begin a
serious discussion of the role of design and design thinking
in HCI research. We will continue to evaluate and refine
our model with practitioners and researchers. Additionally,
we hope to formulate some changes to both HCI and
Figure 4. Concept from Philips vision of the future interaction design education that will allow interaction
project. design research to continue to grow in importance.

Apple Guides ACKNOWLEDGEMENTS


This research through design project helps illustrate how We wish to thank our colleagues at Carnegie Mellons
interaction design research can feed back ideas to the HCI Human-Computer Interaction Institute (HCII) and at the
research community. In designing an interface to a School of Design for their participation, insights, and
multimedia database, the design team used theory from patience. In addition, we would like to thank Professor Dan
cognitive psychology to address the real world issue of Sieworiek, the director of the HCII, for his championing
people getting lost in hypertext interfaces [18]. The team and financial support of this work.
chose to use black and white renderings of people dressed REFERENCES
in historic costumes and set in historic contexts. 1. Alexander, C., Ishikawa, S., Silverstein, M., Jacobson,
In evaluating this interface to see if the guide improved the M., Fiksdahl-King, I., Angel, S. A Pattern Language:
navigation of the content, the design team discovered an Towns, Buildings, Construction. Oxford University
unanticipated of effect. Participants no longer viewed the Press, 1977.
content as having the voice of an unbiased encyclopedia. 2. Apple Computer, Inc. Macintosh Human Computer
Instead, they felt the content represented the opinion of the Interface Guidelines, Addison-Wesley Professional,
individual guide. Through dissemination of these evaluation Reading, MA, 1992.
findings, this design through research project helped to 3. Bayazit, N. Investigating Design: A Review of Forty
stimulate new technical research on the underlying Years of Design Research. Design Issues 20, 1 (2004),
technology to produce embodied agents and new behavioral 16-29.
research to understand the effect embodied agents had on
users. 4. Buchanan, R. Design Research and the New Learning.
Design Issues 17, 4 (2001), 3-23.
CONCLUSION
5. Blevis, E., Lim, Y.K., & Stolterman, E. Regarding
This paper has presented two years of iterative design
Software as a Material of Design. Proc. of
efforts to explore and advance methods for interaction
Wonderground, Design Research Society, (2006).
design researchers to make design research contributions
that both integrate with and benefit the HCI research and 6. Chung, Eric, Jason I. Hong, James Lin, Madhu K.
practice communities. The work has resulted in a new Prabaker, James A. Landay, and Alan Liu. Development
model of interaction design research within HCI that allows and Evaluation of Emerging Design Patterns for
design researchers to collaborate on an equal footing with Ubiquitous Computing. Proc. DIS 2004, ACM Press
HCI engineering and behavioral science researchers. In (2004), 233-242.
addition, it provides a set of critical lenses for evaluating 7. Cross, N. Design Research: A Disciplined Conversation.
what constitutes a good interaction design research Design Issues 15, 2 (1999), 5-10.
contribution for researchers following this model.
8. Cross, N. Designerly Ways of Knowing: Design
The model provides five main benefits. First, it allows the Discipline Versus Design Science. Design Issues 17, 3
HCI research community to engage with wicked problems (2001), 49-55.
that cannot be easily addressed through science and
9. Dunne, A., Raby. F. Design Noir: The Secret Life of
engineering methods. Second, it feeds back technology
Electronic Objects. Birkhuser, Basel, Switzerland,
opportunities to the engineers and gaps in behavior theory
2001.
and unexpected behaviors to the behavioral scientists,
motivating new research. Third, it provides a new method 10. Gaver, WW., Bowers, J., Boucher, A., Gellerson, H.,
for transferring knowledge produced in the HCI research to Pennington, S., Schmidt, A., Steed, A., Villars, N.,
the HCI practice community, potentially increasing the Walker, B. The drift table: designing for ludic
likelihood this knowledge will move into products in the

9
engagement. Ext. Abstracts CHI 04, ACM Press 20. Rheinfrank, J., Hartman, W., Wasserman, A., Design for
(2004), 885-900. usability: crafting a strategy for the design of a new
11. Kidd, C.D., Orr, R., Abowd, G.D., Atkeson, C.G., Essa, generation of Xerox copiers, Usability: turning
I.A., MacIntyre, B., Mynatt, E.D., Starner, T., technologies into tools, Oxford University Press, Inc.,
Newstetter, W. The Aware Home: A Living Laboratory New York, NY, 1992.
for Ubiquitous Computing Research, Proc. of the 21. Rittel, H.W.J., Webber, M.M. Dilemmas in a General
Second International Workshop on Cooperative Theory of Planning. Policy Sciences 4, 2 (1973), 155-
Buildings, Integrating Information, Organization, and 66.
Architecture, (1999) 191-198. 22. Schn, D.A. The Reflective Practitioner: How
12. Fallman, D. Design-Oriented Human-Computer professionals think in action. Temple Smith, London,
Interaction. Proc. CHI 2003, ACM Press (2003), 225- 1983.
232. 23. Simon, H.A. The Sciences of the Artificial, MIT Press,
13. Fallman, D. Why Research-Oriented Design Isn't Cambridge, MA, 1969.
Design-Oriented Research. Proc. NordiCHI 2005, Umea 24. van Duyne, D.K., Landay, J.A., Hong, J. I. The Design
Institute of Design Press (2005). of Sites: Principles, Processes, and Patterns for
14. Frayling, C. Research in Art and Design. Royal College Crafting a Customer-Centered Web Experience,
of Art Research Papers 1, 1 (1993),1-5. Addison-Wesley, Reading, MA, 2003.
15. Laurel B. Design Research: Methods and Perspectives. 25. Wolf, T.V., Rode, J.A., Sussman, J., Kellogg, W.A.
MIT Press, Cambridge, MA, 2003. Dispelling Design as the 'Black Art' of Chi. Proc. of
16. Lwgren, J. Applying Design Methodology to Software CHI 2006, ACM Press (2006), 521-530.
Development. Proc. of DIS 1995, ACM Press (1995), 26. Wright, P., Blythe, M., McCarthy, J. User Experience
87-95. and the Idea of Design in HCI. Lecture Notes in
17. Nelson, H.G. and Stoltermann, E. The Design Way: Computer Science, Stephen W. Gilroy and Michael D.
intentional change in an unpredictable world. Harrison eds. Springer, Berlin/Heidelberg (2006). 1-14.
Educational Technology Publications, Englewood 27. Xerox Corporation. Design Guidelines: Industrial
Cliffs, NJ, 2003. design and Human Factors for Reprographics Products
18.Oren, T., Salomon, G., Kreitman, K., Don, A.: Guides: (1985).
Characterizing the Interface. In (Eds: Laurel, B.): The Art 28. Xerox Corporation. Principles for Constructing
of Human-Computer Interface Design, Addison-Wesley Communicative Objects and Object Systems as
(1990) 355-365. Interactive Dialogs (1985).
19. Philips Design, Vision of the Future V&K (1995)